APPLIED PHYSICS LETTERS
VOLUME 82, NUMBER 10
10 MARCH 2003
Nanoscale molecular-switch devices fabricated by imprint lithography
Yong Chen,a) Douglas A. A. Ohlberg, Xuema Li, Duncan R. Stewart,
and R. Stanley Williams
Quantum Science Research, Hewlett-Packard Laboratories, 1501 Page Mill Road, MS1123, Palo Alto,
California 94304
Jan O. Jeppesen,b) Kent A. Nielsen,b) and J. Fraser Stoddart
Department of Chemistry and Biochemistry and The California NanoSystems Institute,
603 Charles E. Young East Drive, University of California, Los Angeles, California 90095-1596
Deirdre L. Olynick and Erik Anderson
MS 02-400, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, California 94720
共Received 30 August 2002; accepted 16 January 2003兲
Nanoscale molecular-electronic devices comprising a single molecular monolayer of bistable
关2兴rotaxanes sandwiched between two 40-nm metal electrodes were fabricated using imprint
lithography. Bistable current–voltage characteristics with high on–off ratios and reversible
switching properties were observed. Such devices may function as basic elements for future
ultradense electronic circuitry. © 2003 American Institute of Physics. 关DOI: 10.1063/1.1559439兴
Molecular electronics offers the tantalizing prospect of
eventually building circuits with critical dimensions of a few
nanometers. Some basic devices utilizing molecules have
been demonstrated, including tunnel junctions with negative
differential resistance,1 rectifiers,2 and electrically configurable switches that have been used in simple electronic
memoryand logic circuits.3,4 A major challenge that remains
is to show that such devices can be fabricated economically
using a process that will scale to circuits with large numbers
of elements while maintaining their desired electronic properties.
Previous nanoscale molecular devices have been fabricated using e-beam lithography,1,4 which is impractical for
commercial applications because of the slow writing speed.
High-energy electron beams can also damage the active molecules in a circuit. In contrast, imprint lithography is a processing technique that can produce sub-10-nm feature sizes,
high throughput, and low cost.5 In addition, imprinting may
also preclude damage to the active molecules in a circuit
from high-energy electrons during e-beam lithography. In
this letter, we describe an imprinting process to fabricate
nanoscale molecular devices6 from an amphiphilic, bistable
关2兴rotaxane, and demonstrate that these devices act as reversible, electrically toggled switches.
The imprinting mold was fabricated into a 100-nm-thick,
thermally grown silcon oxide on a silicon substrate using
electron-beam lithography and reactive-ion etching 共RIE兲
with a Leica VB6 electron beam writer and an Oxford Instruments Plasmalab 100 ICP180 RIE. The SiO2 surface was
patterned and etched to leave raised mesas of 40-nm-wide
nanowires connected by 3-␮m-wide wires on each end to
100-␮m-square pads. The height of each mesa was 80 nm
above the etched surface of the mold. Each mold had 400
a兲
Author to whom correspondence should be addressed; electronic mail:
yongc@hpl.hp.com
b兲
Current address: Department of Chemistry, University of Southern Denmark 共Odense University兲, Campusvej 55, DK-5230, Odense M, Denmark.
such patterns laid out in a geometrical array, to enable a large
number of devices to be fabricated with one imprinting step.
To form the device electrodes, a 100-nm-thick polymethylmethacrylate 共PMMA兲 film with a 495 K molecular weight
was spin-coated onto a 100-nm-thick SiO2 layer on a silicon
substrate. During imprinting, the mold and polymer-coated
device substrate were first heated to 150 °C, which is higher
than the glass transition temperature of PMMA (105 °C).
The mold was then pressed onto the coated substrate with a
homogeneous pressure of 1000 psi to transfer the pattern
from the mold into the PMMA layer. The mold and device
substrate were then cooled to room temperature before detaching the mold from the PMMA. After imprinting, oxygen
RIE was used to remove ⬃10 nm of residual PMMA at the
bottom of the imprinted trenches, and expose the SiO2 surface. Then, 5-nm-titanium 共Ti兲 and 10-nm-platinum 共Pt兲
metal layers were sequentially evaporated onto the substrate.
A final acetone lift-off process removed the unpatterned field
to leave the Ti/Pt nanowires and their micron-scale connections to the contact pads.
The molecule R that we have used in this work is shown
in Fig. 1. The R label refers to a 关2兴rotaxane, meaning that
the molecule consist of two mechanically interlocked molecular components: a dumbbell encircled by a ring.7 A molecular monolayer of the 关2兴rotaxane R was deposited over
the entire device substrate, including imprinted bottom metal
electrodes, using the Langmuir–Blodgett 共LB兲 method 关Fig.
2共a兲兴. During LB film deposition the aqueous subphase was
maintained at pH⬃8.5 by the addition of tris共hydroxymethyl兲aminomethane, to typical concentrations of 10⫺4 M, at
21 °C.
Fabrication of the top electrodes began with the blanket
evaporation of a 7.5-nm-Ti protective layer. The Ti layer was
very reactive with the top functional group of the molecules8
to form a direct electrical contact to the molecules, which
also blocked the further metal penetration into the LB molecular layers. The Ti layer also minimized subsequent damage to the molecules, enabling further organic resist and sol-
0003-6951/2003/82(10)/1610/3/$20.00
1610
© 2003 American Institute of Physics
Downloaded 27 Mar 2003 to 192.6.19.190. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/aplo/aplcr.jsp
Appl. Phys. Lett., Vol. 82, No. 10, 10 March 2003
Chen et al.
1611
FIG. 1. 共Color兲 Molecular structure of the bistable 关2兴rotaxane R used to
form LB monolayers. The molecule has a hydrophobic 共dark green兲 and a
hydrophilic 共light blue兲 stopper. The 关2兴rotaxane has cyclobis共paraquatp-phenylene兲 共dark blue兲 as the ring component and the supporting
counterions are hexafluorophosphate.
vent processing 关Fig. 2共b兲兴. Patterned top electrodes of 5-nm
Ti and then 10-nm Pt were next fabricated with the same
imprinting process just described, using the same mold. For
the top electrodes, the imprinting mold was oriented perpendicular to the bottom electrodes and aligned to ensure that
the top and bottom nanowires crossed 关Fig. 2共c兲兴. Previous
experiments indicated that the 关2兴rotaxane molecules remained stable up to ⬃210 °C, 9 therefore, the molecular
structure should not be changed during the imprinting process. Finally, RIE with CF4 and O2 共4:1兲 gases at a pressure
of 40 mTorr and a power of 200 W was used to remove the
blanket Ti protective layer anisotropically down to the SiO2
FIG. 3. 共Color兲 An atomic force micrograph of a nanoscale cross-point
molecular device with an insert showing the details of the cross point.
layer, but selectively leave the Pt electrodes intact. The molecules and Ti layer under the Pt top electrode were protected.
After RIE, an array of devices with the molecular monolayer
sandwiched between two metal nanowires remained 关Fig.
2共d兲兴.
An atomic force microscope 共AFM兲 image of a crosspoint molecular device fabricated with our imprint lithography process is shown in Fig. 3. In order to achieve nanometer lateral resolution, a carbon nanotube tip 共ProbeMax兲 was
used as the AFM probe. Only the region near the active part
of the device, comprising the two crossed nanowires and
their connections to the microscale wires, is shown in the
images. The nanowires have a measured width of ⬃40 nm,
which is consistent with the 40-nm width of the nanowire
templates in the mold. A high-resolution image of the
crossed electrodes 关Fig. 3 共inset兲兴 shows the active junction
area of ⬃40 nm⫻40 nm.
Ellipsometric analysis of the 关2兴rotaxane monolayers
yielded an average film thickness of 3.5 nm. A static water
contact angle of ⬃90° indicated a hydrophobic surface and
thus dense monolayer coverage. The active device area of
1600 nm2 corresponds to ⬃1100 molecules sandwiched between the two electrodes, as determined from the known
density of molecules on the surface of the Langmuir trough
before transfer to the substrate. In previous studies utilizing
silicon bottom and titanium top electrodes, switchable
bistable current–voltage characteristics were observed in
both microscale and nanoscale devices with the same 关2兴rotaxane R monolayer.4 In our lab, microscale devices with this
molecule sandwiched between titanium top and platinum
bottom electrodes exhibited switching and continuously tunable resistance over a 102 – 105 ⍀ range under current or
voltage control.10
The devices were tested electrically at room temperature
under ambient conditions with direct I – V sweeps using a HP
4155A semiconductor parameter analyzer and a KarlSuss
PSM6 probe station. The measured resistances of the Ti/Pt
nanowires varied in the range from 17 to 25 k⍀, determined
by two-terminal measurements across a single wire. The
nanowire resistances and the probe/metal contacts
(⬍50 ⍀) were still significantly smaller than the molecular
FIG. 2. 共Color兲 Schematic of the procedure used for fabrication of nanoscale
molecular-switch devices by imprint lithography.
Downloaded 27 Mar 2003 to 192.6.19.190. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/aplo/aplcr.jsp
1612
Chen et al.
Appl. Phys. Lett., Vol. 82, No. 10, 10 March 2003
FIG. 4. Electronic characteristic of a nanoscale molecular device. 共a兲 The
resistances of the device (R d ) recorded 共read兲 at 0.2 V after multiple switching cycles with the open 共filled兲 squares representing R d measured after a
negative 共positive兲 writing voltage cycle applied to the device. 共b兲 I – V
curves showing a complete off–on–off device switching cycle. Curves 1–5
are offset by 20 ␮A from each other for clarity.
monolayer resistance, and they have been neglected in the
subsequent device measurements.
Current–voltage characteristics of the molecular monolayer devices were obtained by applying a test voltage (V T )
to the top electrode while grounding the bottom electrode. As
fabricated, the molecular devices usually showed a very high
resistance, ⬎108 ⍀, when measured at V T ⫽0.2 V. This initial high resistance state was stable for 兩 V T 兩 ⬍2 V. Exceeding
these voltage limits usually caused an irreversible transition
to a smaller resistance. For example, the initial resistance of
a typical device measured at 0.2 V was 6.1⫻108 ⍀, shown
as the first point in Fig. 4共a兲. After sweeping the voltage bias
cycle from 0 to ⫹5 V, the resistance subsequently measured
at 0.2 V dropped to 4.3⫻105 ⍀. After this ‘‘burn-in’’ step,
similar to that reported for our microscale devices,10 the
nanoscale molecular devices became reversible switches
with lower cycling voltages from 0 to ⫾2 V. A typical
switching cycle is shown in Fig. 4共b兲 for the same device
shown in Fig. 4共a兲. To compare the resistances after each
writing process, the resistance for a device (R d ) was always
measured 共read兲 at 0.2 V. After the initial burn-in, the molecular device was set in the ‘‘off’’ state, the I – V characteristic measured in the range ⫾0.2 V showed an ohmic response 关curve 1 in Fig. 4共b兲兴 with R d ⫽8.1⫻106 ⍀. A
positive voltage bias cycle between 0 to 2 V was next applied
to the device, and yielded a counterclockwise I – V hysteresis
loop 关curve 2 in Fig. 4共b兲兴. In this ‘‘on’’ state, the resistance
was R d ⫽4.8⫻105 ⍀ 关curve 3 in Fig. 4共b兲兴. A subsequent
negative voltage bias cycle from 0 to ⫺2 V was then applied
to the device, yielding a clockwise I – V hysteresis 关curve 4
in Fig. 4共b兲兴. The device was again in the off state, with R d
⫽9.2⫻106 ⍀ 关curve 5 in Fig. 4共b兲兴. Figure 4共a兲 shows the
evolution of the measured resistance R d through 95 repetitions of this switching cycle. The on/off switching events
were repeatable, as shown in Fig. 4共a兲. However, the on/off
ratio of R d for this device decreased gradually after ⬃40
switch cycles.
Statistically, ⬃75% of the 36 devices we tested showed
such reversible switching properties. The rest of the devices
were either ‘‘shorted’’ 共with their resistances ⬍105 ⍀ 共comparable to the nanowire resistance兲, or ‘‘open,’’ with resistances ⬎108 ⍀ that remained unchanged even after a high
burn-in voltage (⬎8 V) was applied. The open devices included some with structural defects, such as broken nanowires, detected by post-measurement scanning electron microscopy 共SEM兲. However, some ‘‘defect-free’’ devices as
determined by SEM still had an open electrical character,
perhaps due to a contact problem between the metal electrodes and molecular monolayer. For a device that switched
reversibly, the positive threshold voltage to turn it on ranged
from 0.5 V to 3 V; the negative threshold voltage to switch it
off ranged from ⫺0.5 to ⫺3.5 V. Any voltage bias 兩 V 兩
⬍0.5 V applied to the devices did not change their resistance
state. Both the positive and negative switch threshold voltages varied between devices, and from one switch cycle to
another for a given device. A voltage bias V⬎ 兩 3 V兩 caused
some devices to irreversibly short. At the beginning of
switch cycling, the on/off resistance ratios ranged from 2 to
104 for different devices. The ratio typically decayed below 2
and gradually approached 1 after a few to several hundred
cycles for different devices; several devices measured after
an interval of two months retained the resistance of the state
into which they had last been set.
The authors gratefully acknowledge H. Wiersma, T. Ha,
P. Beck, and S.-H. Leung for experimental assistance, T. I.
Kamins, P. J. Kuekes, Y. Luo, C. P. Collier, and J. R. Heath
for valuable discussions, and the Defense Advanced Research Projects Agency in the United States and by the Carlsbergfondet in Denmark for partial support.
1
J. Chen, M. A. Reed, A. M. Rawlett, and J. M. Tour, Science 286, 1550
共1999兲.
2
R. M. Metzger, Acc. Chem. Res. 32, 950 共1999兲.
3
C. P. Collier, E. W. Wong, M. Belohradský, F. M. Raymo, J. F. Stoddart, P.
J. Kuekes, R. S. Williams, and J. R. Heath, Science 285, 391 共1999兲.
4
Y. Luo, C. P. Collier, J. O. Jeppesen, K. A. Nielsen, E. DeIonno, G. Ho, J.
Perkins, H. Tseng, T. Yamamoto, J. F. Stoddart, and J. R. Heath, Chem.
Phys. Chem 3, 519 共2002兲.
5
S. Y. Chou, P. R. Krauss, and P. J. Renstrom, Science 272, 85 共1996兲.
6
Y. Chen, in U.S. Patent No. 6,432,740 共2002兲.
7
A. R. Pease, J. O. Jeppesen, J. F. Stoddart, Y. Luo, C. P. Collier, and J. R.
Heath, Acc. Chem. Res. 34, 433 共2001兲.
8
K. Konstadinidis, P. Zhang, R. L. Opila, and D. L. Allara, Surf. Sci. 338,
300 共1995兲.
9
J. O. Jeppesen, K. A. Nielsen, J. Perkins, S. A. Vignon, A. D. Fabio, R.
Ballardini, M. T. Gandolfi, M. Venturi, V. Balzani, J. Becher, and J. F.
Stoddart 共unpublished兲.
10
D. R. Stewart, D. A. A. Ohlberg, P. Beck, Y. Chen, R. S. Williams, J. O.
Jeppesen, K. A. Nielsen, and J. F. Stoddart, Appl. Phys. Lett. 共in press兲.
Downloaded 27 Mar 2003 to 192.6.19.190. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/aplo/aplcr.jsp